用于SAR图像小目标舰船检测的改进SSD算法

doi:10.3969/j.issn.1001-506X.2020.05.08

摘要/Abstract

摘要：

针对合成孔径雷达(synthetic aperture radar, SAR)图像小目标舰船检测中存在的检测率较低、虚警较高等缺点,提出了用于SAR图像小目标舰船检测的改进单步多框检测(single shot multibox detector，SSD)算法。首先,制作了一个专门用于SAR图像小目标舰船检测的数据集,在SSD目标检测算法的基础上,提出了迁移学习、浅层特征增强和数据增广3个方面的改进;利用性能更好的ResNet50作为特征提取结构,在浅层特征增强网络结构中采用了inception模块的分支结构,同时使用了空洞卷积扩大特征图的视觉感受野,增强了网络对小尺寸舰船目标的适应性;最后在数据集上进行了多组对比分析实验,实验结果表明所提方法相比于原始的SSD,平均准确率提高了5.4%,并且对SAR小目标舰船的漏检和误报明显减少。

关键词: 目标检测, 卷积神经网络, 迁移学习, 浅层特征增强, 空洞卷积

Abstract:

Aiming at the disadvantages of low detection rate and high false alarm in small-sized ship detection for synthetic aperture radar (SAR) images, this paper proposes a small-sized SAR ship detection algorithm based on convolutional neural network. Firstly, a data set specially designed for small-sized SAR ship detection is constructed. Secondly, based on the single shot multibox detector (SSD) detection algorithm, improvements in transfer learning, bottom feature enhancement, and data augmentation are proposed. Using ResNet50 with better performance as the feature extraction structure, according to the basic principle of the inception module and the dilated convolution to expand the visual receptive field in the bottom feature enhancement, the algorithm enhances the adaptability of the network to small-sized ship targets. In the dataset of this paper, several sets of comparative analysis experiments are carried out. The experimental results show that the proposed method improves the average accuracy by 5.4% compared with the original SSD, and the missing detection and false alarms of small-sized SAR ships are obviously decreased.

Key words: object detection, convolutional neural network(CNN), transfer learning, bottom feature enhancement, dilated convolution

中图分类号:

TP751

苏娟, 杨龙, 黄华, 金国栋. 用于SAR图像小目标舰船检测的改进SSD算法[J]. 系统工程与电子技术, 2020, 42(5): 1026-1034.

Juan SU, Long YANG, Hua HUANG, Guodong JIN. Improved SSD algorithm for small-sized SAR ship detection[J]. Systems Engineering and Electronics, 2020, 42(5): 1026-1034.

图/表 13

图1

表1

图2

图3

图4

图5

图6

图7

图8

表2

表3

表4

图9

参考文献 32

1	HE J L , WANG Y H , LIU H W , et al. A novel automatic PolSAR ship detection method based on superpixel-level local information measurement[J]. IEEE Geoscience & Remote Sensing Letters, 2018, 15 (3): 384- 388.
2	华夏, 王新晴, 王东, 等. 基于改进SSD的交通大场景多目标检测[J]. 光学学报, 2018, 38 (12): 1215003.
	HUA X , WANG X Q , WANG D , et al. Multi-objective detection of traffic based on improved SSD[J]. Acta Optica Sinica, 2018, 38 (12): 1215003.
3	BENGIO Y , COURVILLE A , VINCENT P . Representation learning: a review and new perspectives[J]. IEEE Trans.on Pattern Analysis and Machine Intelligence, 2013, 35 (8): 1798- 1828.
4	谢江荣, 李范鸣, 卫红, 等. 用于空中红外目标检测的增强SSD方法[J]. 光学学报, 2019, 39 (6): 0615001.
	XIE J R , LI F M , WEI H , et al. Enhancement of SSD for aerial infrared target detection[J]. Acta Optica Sinica, 2019, 39 (6): 0615001.
5	郭智, 宋萍, 张义, 等. 基于深度卷积神经网络的遥感图像飞机目标检测方法[J]. 电子与信息学报, 2018, 40 (11): 2684- 2690.
	GUO Z , SONG P , ZHANG Y , et al. Aircraft detection method based on deep convolutional neural network for remote sensing images[J]. Journal of Electronics & Information Technology, 2018, 40 (11): 2684- 2690.
6	朱明明, 许悦雷, 马时平, 等. 基于特征融合与软判决的遥感图像飞机检测[J]. 光学学报, 2019, 39 (2): 71- 77.
	ZHU M M , XU Y L , MA S P , et al. Airplane detection based on feature fusion and soft decision in remote sensing images[J]. Acta Optica Sinica, 2019, 39 (2): 71- 77.
7	刘星, 陈坚, 杨东方, 等. 场景耦合的空对地多任务遥感影像智能检测算法[J]. 光学学报, 2018, 38 (12): 1215008.
	LIU X , CHEN J , YANG D F , et al. Scene-coupled intelligent multi-task detection algorithm for air-to-ground remote sensing image[J]. Acta Optica Sinica, 2018, 38 (12): 1215008.
8	张庆春, 佟国峰, 李勇, 等. 基于多特征融合和软投票的遥感图像河流检测[J]. 光学学报, 2018, 38 (6): 320- 326.
	ZHANG Q C , TONG G F , LI Y , et al. Airplane detection based on feature fusion and soft decision in remote sensing images[J]. Acta Optica Sinica, 2018, 38 (6): 320- 326.
9	GIRSHICK R. Fast R-CNN[C]//Proc.of the IEEE International Conference on Computer Vision, 2015: 1440-1448.
10	REN S , HE K , GIRSHICK R , et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Trans.on Pattern Analysis & Machine Intelligence, 2015, 39 (6): 1137- 1149.
11	DAI J F, LI Y, HE K M, et al. R-FCN: object detection via region-based fully convolutional networks[C]//Proc.of the Advances in Neural Information Processing Systems, 2016: 379-387.
12	HE K, GKIOXARI G, DOLLAR P, et al. Mask R-CNN[C]//Proc.of the IEEE International Conference on Computer Vision, 2017: 2961-2969.
13	LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot multibox detector[C]//Proc.of the European Conference on Computer Vision, 2016: 21-37.
14	LI Z , ZHOU F Q . FSSD: feature fusion single shot multibox detector[J]. Computer Vision and Pattern Recognition, 2018, 36 (7): 356- 366.
15	FU C Y, LIU W, RANGA A, et al. DSSD: Deconvolutional single shot detector[J]. [EB/OL].[2019-10-19]. https://arxiv.org/abs/1701.06659.
16	ZHANG S F, WEN L Y, BIAN X, et al. Single-shot refinement neural network for object detection[C]//Proc.of the IEEE Conference on Computer Vision and Pattern Recognition, 2018: 4203-4212.
17	LIU S T, HUANG D. Receptive field block net for accurate and fast object detection[C]//Proc.of the European Conference on Computer Vision, 2018: 385-400.
18	LIN T Y , GOYAL P , GIRSHICK R , et al. Focal loss for dense object detection[J]. IEEE Trans.on Pattern Analysis & Machine Intelligence, 2017, 2999- 3007.
19	ZHOU P, NI B B, GENG C, et al. Scale-transferrable object detection[C]//Proc.of the IEEE Conference on Computer Vision and Pattern Recognition, 2018: 528-537.
20	REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//Proc.of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 779-788.
21	REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]//Proc.of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 7263-7271.
22	REDMON J, FARHADI A. Yolov3: an incremental improvement[EB/OL]. [2019-10-19]. https://arxiv.org/abs/1804.02767.
23	KANG M , JI K F , LENG X G , et al. Contextual region-based convolutional neural network with multilayer fusion for SAR ship detection[J]. Remote Sensing, 2017, 9 (8): 860.
24	李健伟, 曲长文, 彭书娟, 等. 基于生成对抗网络和线上难例挖掘的SAR图像舰船目标检测[J]. 电子与信息学报, 2019, 143- 149.
	LI J J , QU C W , PENG S J , et al. Ship detection in SAR images based on generative adversarial network and online hard examples mining[J]. Journal of Electronics and Information Technology, 2019, 143- 149.
25	GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Gene-rative adversarial nets[C]//Proc.of the Advances in Neural Information Processing Systems. Montreal, 2014, 2672-2680.
26	JIAO J , ZHANG Y , SUN H , et al. A densely connected end-to-end neural network for multiscale and multiscene SAR ship detection[J]. IEEE Access, 2018, 6 (1): 20881- 20892.
27	胡昌华, 陈辰, 何川, 等. 基于深度卷积神经网络的SAR图像舰船小目标检测[J]. 中国惯性技术学报, 2019, (3): 397- 405, 414.
	HU C H , CHEN C , HE C , et al. Detection and recognition of SAR small ship objects using deep neural network[J]. Journal of Chinese Inertial Technology, 2019, (3): 397- 405, 414.
28	李健伟, 曲长文, 彭书娟. SAR图像舰船目标联合检测与方向估计[J]. 武汉大学学报·信息科学版, 2019, 44 (6): 901- 907.
	LI J W , QU C W , PENG S J . A joint SAR ship detection and azimuth estimation method[J]. Geomatics and Information Science of WuHan University, 2019, 44 (6): 901- 907.
29	魏松杰, 蒋鹏飞, 袁秋壮, 等. 深度神经网络下的SAR舰船目标检测与区分模型[J]. 西北工业大学学报, 2019, 37 (3): 587- 593.
	WEI S J , JIANG P F , YUAN Q Z , et al. Detection and recognition of sar small ship objects using deep neural network[J]. Journal of Northwestern Polytechnical University, 2019, 37 (3): 587- 593.
30	SZEGEDY C, VANHOUCKE V, IOFFE S, et al. Rethinking the inception architecture for computer vision[C]//Proc.of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 2818-2826.
31	IOFFE S, SZEGEDY C. Batch normalization: accelerating deep network training by reducing internal covariate shift[C]//Proc.of the International Conference on Machine Learning, 2015.
32	LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft COCO: common objects in context[C]//Proc.of the European Conference on Computer Vision, 2014: 740-755.

卷积层	特征图尺寸/像素	感受野大小/像素
Conv4_3	38×38	92×92
Conv7	19×19	276×276
Conv8_2	10×10	340×340
Conv9_2	5×5	468×468
Conv10_2	3×3	724×724
Conv11_2	1×1	980×980

算法	网络结构	图片尺寸/像素	AP/%	FPS
SSD	VGG16	300×300	82.7	45
SSD	ResNet50	300×300	83.4	40
FSSD	VGG16	300×300	83.9	36
FSSD	ResNet50	300×300	84.8	29
本文算法	ResNet50	300×300	88.1	33

算法	网络结构	图片尺寸/像素	AP/%	FPS
Faster R-CNN	VGG16	600×1 000	86.3	10
本文算法	ResNet50	300×300	88.1	33

[1]	宋爽, 张悦, 张琳娜, 岑翼刚, 李浥东. 基于深度学习的轻量化目标检测算法[J]. 系统工程与电子技术, 2022, 44(9): 2716-2725.
[2]	肖宇, 邓正宏, 张展. 基于双阶段互信息准则的多目标检测波形设计[J]. 系统工程与电子技术, 2022, 44(9): 2736-2742.
[3]	杨立儒, 刘永祥, 杨威. 基于迁移学习的雷达杂波幅度统计模型选择[J]. 系统工程与电子技术, 2022, 44(8): 2457-2467.
[4]	王彩云, 吴钇达, 王佳宁, 马璐, 赵焕玥. 基于改进的CNN和数据增强的SAR目标识别[J]. 系统工程与电子技术, 2022, 44(8): 2483-2487.
[5]	刘祥, 黄天耀, 刘一民. 频率捷变雷达的扩展目标检测[J]. 系统工程与电子技术, 2022, 44(6): 1833-1838.
[6]	孙晶明, 虞盛康, 孙俊. 基于元学习的雷达小样本目标识别方法及改进[J]. 系统工程与电子技术, 2022, 44(6): 1839-1845.
[7]	韦娟, 杨皇卫, 宁方立. 基于NMF与CNN联合优化的声学场景分类[J]. 系统工程与电子技术, 2022, 44(5): 1433-1438.
[8]	赵晓枫, 徐叶斌, 吴飞, 牛家辉, 蔡伟, 张志利. 基于全局感知机制的地面红外目标检测方法[J]. 系统工程与电子技术, 2022, 44(5): 1461-1467.
[9]	魏文晓, 刘洁瑜, 沈强, 李成. 基于人眼视点图的特征融合小目标检测算法[J]. 系统工程与电子技术, 2022, 44(4): 1120-1127.
[10]	陈冬, 句彦伟. 基于语义分割实现的SAR图像舰船目标检测[J]. 系统工程与电子技术, 2022, 44(4): 1195-1201.
[11]	周晓玲, 张朝霞, 鲁雅, 王倩, 王琨琨. 基于改进R-FCN的SAR图像识别[J]. 系统工程与电子技术, 2022, 44(4): 1202-1209.
[12]	方伟, 王玉, 闫文君, 林冲. 基于神经网络的符号化飞行动作识别[J]. 系统工程与电子技术, 2022, 44(3): 737-745.
[13]	李洪瑶, 李小强, 韩心中, 谢学立, 席建祥. 基于决策融合的多无人机协同目标检测识别算法[J]. 系统工程与电子技术, 2022, 44(3): 746-754.
[14]	孙晶明, 虞盛康, 孙俊. 基于深度学习的HRRP识别姿态敏感性分析[J]. 系统工程与电子技术, 2022, 44(3): 802-807.
[15]	何立, 沈亮, 李辉, 王壮, 唐文泉. 强化学习中的策略重用: 研究进展[J]. 系统工程与电子技术, 2022, 44(3): 884-899.